2.CDMA, DECORRELATING DETECTOR, CONVENTIONAL DETECTOR, FUZZY ADAPTIVE DETECTOR, MULTIUSER DETECTION (MUD), LMS, RLS, ANFIS, FUZZY LOGIC SYSTEM, İbrahim ŞİMŞEK, 2004

Keywords : CDMA, Decorrelating detector, Conventional detector, Fuzzy Adaptive detector, Multiuser detection (MUD), LMS, RLS, ANFIS, Fuzzy logic system

Spread spectrum systems have been widely used in the military for decades. Nowadays spread spectrum systems such as Code Division Multiple Access (CDMA) has taken on a significant role in commercial communication. In CDMA systems, multiple users share common channel simultaneously and the received signal consists of signals from multiple users. A matched filter bank has been used to detect CDMA signals but the conventional matched filter bank detector cannot avoid multiple access interference (MAI) due to the cross correlation existing between spreading codes. In addition, because each user’s received signal power has different value, when interfering signals are stronger than desired signal, the conventional matched filter bank detector suffers from severe near-far problem.

To solve the near-far problem and to reduce the MAI, a decorrelating detector was proposed. The decorrelating detector can decouple matched filter outputs by multiplying the received signal with the inverse of cross correlation matrix between spreading codes. In the absence of background noise, the decorrelating detector achieves perfect demodulation unlike the matched filter bank. The advantage of the decorrelating detector is that it does not require knowledge of the received signal amplitudes.

However, the disadvantages of the decorrelating detector are that the computations needed to inverse cross correlation matrix are difficult to perform in real time for many user systems and that at low SNR’s the decorrelating detector can perform even worse than the conventional detector.

In this study, a RLS adaptive fuzzy multiuser detector for CDMA systems is proposed. The enhanced detection capabilities enable to increase performance and capacity of CDMA systems. Although, fuzzy systems have widely been used in many applications in control systems recently, the application of fuzzy systems to mobile communications has been limited.

In the proposed system, fuzzy adaptive multiuser detectors are constructed as a set of fuzzy IF-THEN rules which change adaptively using LMS and RLS algorithms during a short training period, to minimize the bit error rate. Detection results are compared to the conventional detector and the decorrelating detector. The RLS adaptive fuzzy multiuser detector gives the best BER performance clearly demonstrating the superiority of the proposed system.